Summary

  • Indonesia’s draft assigns village-level green, yellow and red zones to different mobile-network build roles, with state entry in the least commercial areas.
  • The state network has no calendar sunset: a carrier must provide equal or better coverage, availability and quality before public provision ends, and must then use the state infrastructure and pay its operating costs.

A map can look like a neutral record of where a network exists. In Indonesia’s draft mobile-broadband plan, it is closer to an allocation instrument. A village or urban ward’s zone would shape who is expected to invest, which technical facilitation may apply and whether the communications ministry can build infrastructure. The label would also determine the service threshold used to assess a part of the country’s 2029 connectivity ambition.

The Ministry of Communication and Digital published the draft for consultation on 23 September 2026. It divides desa and kelurahan into green, yellow and red zones. Green means the area is considered commercially viable. Yellow means an operator is expected to build but may need facilitation. Red means the area is not considered commercially viable and calls for a joint plan between the ministry and cellular network operators. Those are proposed categories, not evidence that a particular village currently lacks service or that a build is funded.

The draft’s classification inputs mix demand and supply. On the demand side are population, households, electricity users, road access, business activity, tourism, education and health. The supply side includes fibre-optic network providers, microwave links, the number and density of mobile sites, and signal coverage measured against population, land area and settlements. This is more revealing than a simple “urban versus rural” map: the plan would combine the economic case for a location with the infrastructure already present.

Yet Article 5 names the factors without stating how they are weighted or where the boundaries between zones sit. The 983-page consultation file includes long annex lists of villages and wards by zone, but a list does not show why one place crossed a threshold while another did not. That leaves an important difference between publishing a map and publishing the method that made it. The draft could be backed by a detailed working dataset; the public text reviewed here does not disclose a scoring formula.

The zone then changes the build arrangement. In green areas, cellular operators provide infrastructure. In yellow areas they remain the lead builder, with proposed relief that includes exceptions to minimum separation rules for microwave-link radio stations and permission to use radio local area network spectrum outside the ordinary operating conditions. The ministry could step in temporarily in a yellow area when local demand exists, but only through joint planning agreed with an operator. Red areas are planned jointly from the outset. The ministry may serve a different part of the same village or ward than the operator.

This is not simply a public subsidy paid once and forgotten. Article 8 says government provision in a red area is temporary until a mobile operator supplies infrastructure with the same or broader coverage, the same or better service availability, and the same or better quality. It sets no date by which that must happen. Article 9 then requires the operator to use government infrastructure already provided in the same area and places the operating cost on the operator. The arrangement is therefore a conditional asset transition, not an automatic public exit when a commercial licence or tower appears.

The draft also gives a measurable, though incomplete, meaning to coverage. Under Article 15, an area counts as covered when signal coverage reaches at least 80% of either the village’s land area or its population at a signal level no weaker than -105 dBm, aggregated across the spectrum in use. “Area or population” is a choice of denominator; it does not require both tests to pass. Article 16 proposes average-user speed floors of 18 Mbps in yellow zones and 3 Mbps in red zones. Green zones follow the minister’s strategic plan, and the Director-General is to set the calculation procedure.

Those local definitions are not the same as the national goal cited by Komdigi: mobile-broadband coverage of 98% of the population by 2029. The government target is population-based and countrywide; the draft’s test asks whether one village or ward reaches an 80% threshold by area or population. The two measures can inform one another, but neither is a substitute for the other. Nor does a signal threshold or average speed alone establish affordability, indoor performance, busy-hour capacity, device access or actual use.

The map is designed to persist until a formal update. If a village is divided or shifted administratively, Article 12 keeps the existing zone. If location references conflict, Article 13 points to the Directorate-General’s map. Article 17 allows an update no earlier than one year after newer village-potential data is published; an area thought to need adjustment remains under its current treatment until a new plan is issued. That may protect the rollout from constant administrative churn, but it also means a new road, school, clinic, population change or private network does not by itself change the formal zone.

The consultation closed on 5 October. The text remains a draft in the sources reviewed, with placeholders for its regulation number and promulgation date. Its 98% target is an objective, not a measured result; no village-level build, carrier commitment or handoff is established by the proposal itself.

Heng Lu Note 64 is useful here only as an editorial question about what a common plan needs to specify and what should remain responsive to later evidence. Its design doctrine concerns Internet coordination systems, not Indonesian telecom law. The relevant test for this draft is more concrete: can residents, operators and public agencies see how a place was zoned, what a transition requires, and when outdated inputs can be challenged?

Sources: Komdigi’s consultation notice, the 983-page draft regulation and annexes, Komdigi’s 98% target announcement, and Heng Lu Note 64.